Kim Chwee Daniel
Tan
*a,
Keith S.
Taber
b,
Yong Qiang
Liew
c and
Kay Liang Alan
Teo
d
aNational Institute of Education, Nanyang Technological University, Singapore. E-mail: daniel.tan@nie.edu.sg
bFaculty of Education, University of Cambridge, UK
cMillennia Institute, Singapore
dHwa Chong Institution, Singapore
First published on 13th February 2019
The internet is prevalent in society today, and user-friendly web-based productivity tools are readily available for developing diagnostic instruments. This study sought to determine the affordances of a web-based diagnostic instrument on ionisation energy (wIEDI) based on the pen-and-paper version, the Ionisation Energy Diagnostic Instrument (IEDI) that was previously developed and reported on. The Google Forms platform was used to develop the wIEDI and it allowed a degree of personalisation such that specific second-tier options are offered in response to the student's choice of answer in the first tier. Students could choose one or more reasons in the second tier or supply their own reasons, and they were asked to indicate their confidence in their choice of answer–reason combinations. The wIEDI was administered to 274 A-level students (257 Grade 11 and 17 Grade 12), and answer–reason combinations indicating alternative conceptions were highlighted only if 5% or more students expressed confidence in them in the third-tier confidence measure. The results showed that all the possible alternative conceptions of ionisation energy reported in the previous study were also identified in the present study. Additional alternative conceptions were indicated as new reasons had to be developed for many items in the wIEDI to ensure that there were sufficient reasons for each first-tier response, and students were allowed to choose more than one reason for their answer. The wIEDI better facilitated responses reflecting the consistency of the use of specific ideas in student thinking and provided direct evidence of students’ possible manifold conceptions and thinking within each question as well as across a range of questions. It also allowed easy collation of the comments students typed in response to the ‘Others’ and ‘I do not know the answer’ options. Thus, the study makes a case for researchers and teachers using such technology in the diagnostic assessment of students.
Thus, it is not surprising that research has shown that students have difficulty in understanding the concepts involved in ionisation energy (Taber, 1999, 2003; Tan et al., 2005a, 2005b; Taber and Tan, 2007). When asked to explain the aspects of patterns in ionisation energies, many A-level students (16 to 18 years old) used the octet rule/full shell framework to state that atoms ‘want’ to obtain full outermost shells or octets of electrons in the outermost shell, and once they are achieved, the atoms are ‘reluctant’ to lose these electronic configurations; the students seemed to think that the sodium ion already has an octet configuration so it will not gain an electron to reform the sodium atom (Taber, 2013a). The conservation of force thinking was also commonly seen in the students’ explanations – that the attractive force arising from a charged entity will be shared amongst oppositely-charged entities. For example, students thought that if an electron left a sodium atom, the remaining 10 electrons would share the attraction from the 11 protons in the nucleus that had previously been shared among the 11 electrons. Using a two-tier multiple choice diagnostic instrument where students were required to choose an answer from the first-tier options and a reason for the answer from the second-tier options, Tan and his colleagues found that students also used the stable fully-filled or half-filled subshell conceptions (these configurations are considered to be ‘stable’ in a similar sense to ‘stable’ octets) and relation-based reasoning (considering only one or two factors instead of all the factors involved) (Driver et al., 1996) to justify their answers on the ionisation energy trend across Periods 2 and 3 of the Periodic Table. The students seemed to have difficulty in applying basic electrostatic principles that they learned in physics, if they were also taking A-level physics in the first instance, to explain the interactions between the nucleus and electrons in an atom/ion and between electrons in orbitals, and how these interactions influenced the trends of ionisation energy (Taber, 1999; Cann, 2000).
In addition, studies on students’ understanding of ionisation energy showed that students’ responses to similar items were inconsistent (Taber, 1999, 2003; Tan et al., 2005a, 2005b; Taber and Tan, 2007), with some students using “both the correct concepts and alternative modes of thinking, applying a different way of thinking to different items” (Taber and Tan, 2007, p. 386). For example, Tan et al. (2005b) reported that several students who applied the appropriate electrostatics concepts to correctly choose the answer and reason for an item on the re-formation of the sodium atom did not use similar concepts to justify their answer for an item on the stability of the species involved in the ionisation process; instead, they used the octet rule/full shell framework. Gilbert and Watts (1983) did strike a cautionary note when they stated that “it may well be that an individual's conceptions make use of several frameworks” (p. 86), and some evidence of this has been found when a student's thinking is explored in detail (Taber, 2000). Palmer (1999) was in agreement when he suggested that “different conceptions can be brought into play in response to different problem contexts” (p. 639). It has also been proposed that students can have stable conceptual resources and/or p-prims, which are interpretations of perceptual data (diSessa, 1993, 2018), that they find useful and relevant (although this may refer to pre-conscious processing) in constructing knowledge in response to a particular context at a point in time, but they do not apply these elements consistently (Taber, 2006; Taber and Tan, 2007). This ‘Knowledge in Pieces’ perspective (diSessa, 2018) is helpful in explaining why students can give the correct answers, and then show that they have alternative conceptions, even a range of different ones, for other similar questions. This discussion assumes that the elicited student's response reflects his/her “deep-rooted well-considered and significant beliefs about the world” and not “just ‘throw-away’ suggestion(s)” (Taber, 2006, p. 152) or guessing in response to questions asked during interviews or tests. Asking students to report their levels of confidence in their responses can provide researchers or teachers with more information in drawing such inferences.
When two-tier multiple choice instruments are used, another possible explanation for a student's inconsistent responses to questions involving very similar concepts is that he/she is forced to choose one reason for his/her answer even if the other reasons presented are also attractive to him/her. If the student has more than one conception for a given concept (Palmer, 1999; Taber, 2013b) or has several applicable p-prims which he/she can use to construct answers (diSessa, 1993, 2018), then allowing the student to choose more than one reason can enable him/her to display all the relevant conceptions or p-prims that he/she has for items assessing similar concepts. The researcher can track these reasons across related items instead of only one reason per item, and this could help increase the consistency of the student's responses across the related items.
Allowing several responses may seem inconsistent with the common assessment practice of offering one correct response with several incorrect distractors for an objective item. However, arguably, authentic assessment items should sometimes admit more than one correct response given that in many contexts of scientific interest (such as comparing ionisation energies between elements) there may be several relevant factors at work, and given that often the same phenomenon may have complementary explanations at several different levels of analysis (for example, in chemistry, using complementary macroscopic concepts and submicroscopic models).
Hasan et al. (1999) suggested the use of the Certainty of Response Index (CRI) with the items on diagnostic tests to differentiate between students’ lack of knowledge (and resorting to guessing) and possible alternative conceptions. In this protocol, respondents are requested to indicate how confident they are, on a six-point scale (0–5), of their knowledge required to answer a given question. So, if the degree of certainty is low (CRI of 0–2), it will be assumed that the answer is mainly due to guesswork, and if the degree of certainty is high (CRI of 3–5) and the answer is wrong, it will be assumed that the student has a possible alternative conception. Researchers have incorporated confidence testing into two-tier multiple choice diagnostic instruments to produce three-tier tests (Caleon and Subramaniam, 2010a; Brandriet and Bretz, 2014) and four-tier tests (Caleon and Subramaniam, 2010b; Sreenivasulu and Subramaniam, 2013; Lin, 2016; Yan and Subramaniam, 2018) to determine the students’ confidence of their answer–reason combinations or their confidence of their choices in each tier separately, and hence, whether students have genuine or spurious alternative conceptions (Caleon and Subramaniam, 2010a). Whether such an additional tier, with the extra demand on respondents, is indicated may in part depend on the purpose of an instrument. When used for classroom diagnostic assessment by teachers, the outcomes of administering such an instrument can (and should) be the basis of classroom follow-up where students can discuss their ideas further as part of a dialogue to help shift them towards thinking with the relevant scientific models.
It is easy to create two-tier, three-tier or four-tier multiple choice diagnostic instruments using web-based productivity and assessment tools, such as Google Forms, without much technical expertise (Tan and Koh, 2008). Web-based instruments created using Google Forms have the advantages of being easily accessible to teachers and students, being readily administered to students, and allowing students’ answers to be collated automatically; the data can be downloaded as Microsoft Excel documents, allowing further processing and analyses of students’ responses to be undertaken. The Google Forms platform can inform students of their results and the answer for each item, but providing adaptive feedback, that is, diagnosing students’ difficulties if they get the items wrong and responding to these difficulties to help them attain the required learning (Devedžić, 2006), requires much more content and item development such as would be possible on platforms such as the WileyPlus ORION Adaptive Practice (Wiley, 2018).
It may be challenging to allow students to choose more than one reason in pen-and-paper tests as optical mark readers, when used to facilitate the collation of students’ answers, might not be able to process students’ choice of more than one reason per item. However, this function can be facilitated by Google Forms using a checklist format for the second-tier reasons. In addition, Google Forms can also address the issue of students treating the options in the two tiers as if they are independent of each other, allowing answer–reason combinations which are seemingly illogical and difficult to interpret (Griffard and Wandersee, 2001; Tan et al., 2002; Loh et al., 2014). This is achieved by allowing a specific set of reasons to be presented to the students based on the answer that they selected. It has to be acknowledged that a response combination that makes no sense to the researcher or teacher can be an arbitrary combination made by a student, a mistaken tick in the wrong box, or a response that makes at least as much sense to the student at that time as the other available permutations, including the orthodox ones.
Thus, the aim of the study was to determine the affordances of a web-based two-tier multiple choice diagnostic instrument on ionisation energy based on the pen-and-paper version developed in a previous study. It attempts to make a case for why researchers and teachers should consider using computer-based or web-based diagnostic instruments for the formative assessment of students.
New options were included in the second tiers so that there were at least two options in any set of second-tier reasons (see Fig. 1 for a comparison of item 1 in the original instrument and in the web-based instrument). These additional options were derived from the second-tier options of other related items, as well as less popular responses given in the open-ended versions of the instrument and interviews from the previous study which were not adopted as options for the paper-based IEDI (where all students need to be presented with the same second-tier options regardless of their first-tier choice) as only four to five second-tier options were required for each item in the IEDI. As previously mentioned, Google Forms allowed the researchers to easily format the second tier as a checklist so that the student could choose more than one reason and even write his/her own reason if he/she wished. Again, there may be an issue here if students see the instrument as a test, and are familiar with multiple choice tests where there is always one correct response and a number of incorrect distractors. However, in a topic such as ionisation energy, explanations may involve more than one factor, for example, nuclear charge, ionic radius, the type of orbital (i.e., the value of the azimuthal quantum number) and the type of shell (i.e., the value of the principal quantum number), so students are expected to consider a range of factors at one time. Thus, it is important that students are aware that they can select several options rather than just one response option, as well as write their own reasons if necessary.
There were discussions among the researchers on the number of points to have on the confidence scale and where to place the scale – for the first tier only, the second tier only, both the first and second tiers, or the whole item (answer–reason combination). It was decided that the confidence interval would cover the whole item as the researchers wanted the students to indicate that their confidence in their answer–reason combinations. The analyses of results would focus on answer–reason combinations rather than the answer or reason(s) separately as the specific sets of reason options offered to the students were dependent on the answers chosen by them. Rather than having a multi-point confidence scale and then calculating the mean to differentiate between students’ lack of knowledge (and resorting to guessing) and possible alternative conceptions or understanding, the researchers decided to have two points only, ‘Confident’ or ‘Not confident’, so students need only consider two options. If students did not know the answer to the first tier, they could choose the ‘I do not know the answer’ option rather than guessing an answer. In this case, the confidence level tier would not be offered as it would not serve any meaningful purpose to ask if they were confident or not confident that they did not know the answer.
The teachers had the freedom to select the number of classes/students to administer the instrument to. They forwarded the students involved an email from the first-named author containing a brief description of the study and the address of a website which contained more detailed information. In the website, the parents of the students were requested to give their consent to the participation of their children in the study; parents, of course, could withhold consent and the process would end with a message thanking them for taking time to visit the website. The students were also requested to give their assent to their participation in the study if their parents had given their consent. If the students declined to give assent, the process would also end with a ‘thank you’ message. The sampling method used, relying on personal contacts, and students opting in, prevents us from claiming that the results reported here are representative of the population of A-level chemistry students in Singapore. However, the main purpose of this study was not to survey a population, but to develop and trial a web-based instrument to determine its affordances.
Alternative conception indicated by reason | Item and reason | Percentage of students (n = 274) | Percentage of students in the previous study (n′ = 979) |
---|---|---|---|
Notes: * indicates that the percentage of students was less than 10% in the previous study (Tan et al., 2005a, 2005b). N.A. ** indicates that the option was not present in the previous study. | |||
Octet rule framework (oct) | |||
The sodium ion will not recombine with an electron to reform the sodium atom because the sodium ion has a stable/noble gas configuration, so it will not gain an electron to lose its stability. | Q1 (A2) | 17 | 44 |
The Na(g) atom is a less stable system than the Na+(g) ion and a free electron because the outermost shell of the Na+(g) ion has achieved a stable octet/noble gas configuration. | Q3 (B2) | 36 | 64 |
The second ionisation energy of sodium is greater than its first ionisation energy because removal of the second electron disrupts the stable octet structure of the Na+ ion. | Q4 (A1) | 22 | 16 |
The first ionisation energy of sodium is less than that of magnesium because sodium will achieve a stable octet configuration if its 3s electron is removed. | Q5 (B1) | 7 | 9* |
The first ionisation energy of sodium is less than that of aluminium because sodium will achieve a stable octet configuration if an electron is removed. | Q7 (B1) | 5 | 6* |
Stable fully-filled subshell conception (ffs) | |||
The Na(g) atom is a less stable system than the Na+(g) ion and a free electron because all subshells are filled fully with electrons for Na+(g), whereas the Na atom has one subshell with a lone electron that gives it its instability. | Q3 (B3) | 26 | N.A.** |
The second ionisation energy of sodium is greater than its first ionisation energy because it is more difficult to remove an electron from a fully-filled 2p subshell. | Q4 (A4) | 25 | N.A.** |
The first ionisation energy of sodium is less than that of magnesium because magnesium has a fully-filled 3s subshell which gives it stability as paired electrons are more stable and harder to remove. | Q5 (B2) | 11 | 13 |
The first ionisation energy of magnesium is greater than that of aluminium because removal of an electron will disrupt the stable completely filled 3s subshell of magnesium. | Q6 (A1) | 5 | 6* |
Stable half-filled subshell conception (hfs) | |||
The first ionisation energy of silicon is less than that of phosphorus because the 3p subshell of phosphorus is half-filled; hence it is stable. | Q8 (B1) | 8 | 25 |
The first ionisation energy of phosphorus is greater than that of sulfur because the 3p subshell of phosphorus is half-filled; hence it is stable. | Q9 (A2) | 7 | 20 |
The first ionisation energy of phosphorus is greater than that of sulfur because sulfur needs to lose one electron to have a stable half-filled 3p subshell. | Q9 (A4) | 8 | N.A.** |
Conservation of force conception (cof) | |||
When an electron is removed from the sodium atom, the attraction of the nucleus for the ‘lost’ electron will be redistributed among the remaining electrons in the sodium ion because the number of protons in the nucleus is the same but there is one less electron to attract, so the remaining 10 electrons will experience greater attraction by the nucleus. | Q2 (A1) | 37 | 50 |
When an electron is removed from the sodium atom, the attraction of the nucleus for the ‘lost’ electron will be redistributed among the remaining electrons in the sodium ion because the Na+(g) ion has one less shell compared to the Na(g) atom. | Q2 (A2) | 14 | N.A.** |
When an electron is removed from the sodium atom, the attraction of the nucleus for the ‘lost’ electron will be redistributed among the remaining electrons in the sodium ion because the total force of attraction by the nucleus is always divided equally between the total number of electrons. | Q2 (A3) | 9 | N.A.** |
The Na(g) atom is a less stable system than the Na+(g) ion and a free electron because the average force of attraction on each electron of the Na+(g) ion is greater than that of the Na(g) atom as the attractive force of the nucleus of the Na+(g) is shared among fewer electrons. | Q3 (B1) | 6 | 1* |
The second ionisation energy of sodium is greater than its first ionisation energy because the same number of protons in the Na+(g) ion attract one less electron, so the attraction for the remaining electrons is stronger. | Q4 (A2) | 42 | 18 |
Strongly electropositive sodium only loses electrons (pos) | |||
The sodium ion will not recombine with an electron to reform the sodium atom because sodium is strongly electropositive, so it only loses electrons | Q1 (A1) | 8 | 5* |
In discussing responses to the items, we refer readers to the Appendix where each item is presented with the first-tier response options (denoted by letters) and their associated potential reasons presented (denoted by numbers). We also use some abbreviations to denote the alternative conceptions linked to particular response options:
• (oct) – octet rule framework
• (ffs) – stable fully-filled subshell conception
• (hfs) – stable half-filled subshell conception
• (cof) – conservation of force conception
• (pos) – strongly electropositive sodium only loses electrons
So, for example, for item 1 response A2 (oct) indicates a confident response option of A (on the first tier) and 2 (on the second tier), which is linked to thinking using the octet rule framework. Where responses in two or more items related to an alternative conception (e.g. the octet rule framework in items 1 and 3) are discussed, these responses may be grouped together {e.g. [Q1 (A2), Q3 (B2)] (oct)}.
The percentages in Table 1 include students who had chosen a particular option in all possible combinations as they could choose more than one reason for an answer. For example, 17% of the total number of students confidently chose the ‘octet rule framework’ reason A2 (oct) in item 1. This included students who chose option A2 only, as well as the combinations (A1, A2) only, (A1, A2, A3) only, (A1, A2, A4) only, (A2, A3) only, (A2, A3, A4) only and (A2, A4) only. There is double counting in the sense that a few of the above several combinations [(A1, A2), (A1, A2, A3), (A1, A2, A3, A4)] are also included in the computation of the percentage of students who confidently chose A1 (pos), the ‘strongly electropositive sodium only loses electrons’ option in item 1. Thus, the total percentage of students who confidently chose all possible answer–reason combinations in an item could exceed 100%.
Alternative conception | Items cross-tabulated | Percentage of students (n = 274) |
---|---|---|
Octet rule framework (oct) | Q1 (A2), Q3 (B2) | 8 |
Q1 (A2), Q4 (A1) | 5 | |
Q3 (B2), Q4 (A1) | 10 | |
Stable fully-filled subshell conception (ffs) | Q3 (B3), Q4 (A4) | 9 |
Conservation of force (cof) | Q2 (A1), Q2 (A2) | 13 |
Q2 (A1), Q2 (A3) | 9 | |
Q2 (A1), Q4 (A2) | 24 | |
Q2 (A2), Q4 (A2) | 8 | |
Q2 (A3), Q4 (A2) | 6 | |
Q2 (A1), Q2 (A2), Q4 (A2) | 8 | |
Q2 (A1), Q2 (A3), Q4 (A2) | 5 |
The stable fully-filled subshell conception was common among students (see Table 1). Reasons B3 (ffs) of item 3 and A4 (ffs) of item 4 were created to ensure that there were sufficient reasons for the related answers in both items. These two reasons were effective distractors, each attracting about 25% of the students who were confident of their choices. The only instance of the ‘stable fully-filled subshell conception’ reasons consistently and confidently chosen by 5% or more students over two items was B3 (ffs) in item 3 and A4 (ffs) in item 4 (9% of 274 students) (see Table 2). Without these two additional reasons created for the wIEDI, there would not have been any response options consistent with the stable fully-filled subshell conception (no instance was highlighted in the previous study). The lack of consistency of students’ choices was also highlighted in the earlier study, and this was attributed to students’ holding manifold conceptions and different conceptions coming into play in response to different contexts (Gilbert and Watts, 1983; Palmer, 1999; Taber 1999, 2000). This explanation seemed to be supported in this study – even though the students were allowed to choose as many reasons as they believed supported the answer that they chose, they did not confidently and consistently choose the stable fully-filled subshell conception across all four items. As previously mentioned, items 3 and 4 focused on the Na+ ion, which has an octet of electrons in its outermost shell as well as fully-filled 2p subshells, so students might have been ‘directed’ to focus on the electronic configuration of the Na+ ion in these items when considering the reasons for their answers, but not in items 5 and 6.
The existence of manifold conceptions was evident in Table 3, where students could be seen choosing many reasons for an answer in the same item, even the correct reason as well as reasons indicating alternative conceptions; providing direct evidence of manifold conceptions is an important affordance of the web-based diagnostic instrument. The percentage for a given cross-tabulation in Table 3 was derived from the total number of students’ responses containing the specific answer–reason combinations. For example, in item 4, 18% of 274 students confidently chose reason A1 (oct) and the correct reason A3* in all combinations (second entry). This figure includes the students who confidently chose (A1, A3*) only as well as (A1, A2, A3*) only, (A1, A2, A3*, A4) only, (A1, A2, A3*, Other) only, (A1, A3*, A4) only, and (A1, A3*, Other) only. A number of students (5%) confidently chose all four reasons: A1 (oct), A2 (cof), A3* and A4 (ffs)! Being allowed to choose more than one reason for an answer in an item (and the students had to indicate that they were confident of their choices) in this study presented direct evidence of students’ manifold conceptions compared to the earlier study where students’ manifold conceptions were inferred through the lack of consistency of their choices across items.
Item | Reasons chosen | Percentage of students (n = 274) |
---|---|---|
1 | A1 (pos), A2 (oct) | 5 |
B1*, B2 (rxn) | 6 | |
2 | A1 (cof), A2 (cof) | 13 |
A1 (cof), A3 (cof) | 9 | |
3 | B2 (oct), B3 (ffs) | 25 |
4 | A1 (oct), A2 (cof) | 9 |
A1 (oct), A3* | 18 | |
A1 (oct), A4 (ffs) | 12 | |
A2 (cof), A3* | 33 | |
A2 (cof), A4 (ffs) | 12 | |
A3*, A4 (ffs) | 21 | |
A1 (oct), A2 (cof), A3* | 9 | |
A1 (oct), A2 (cof), A4 (ffs) | 5 | |
A1 (oct), A3*, A4 (ffs) | 10 | |
A2 (cof), A3*, A4 (ffs) | 11 | |
A1 (oct), A2 (cof), A3*, A4 (ffs) | 5 |
Interestingly, there were 12 instances of 5% or more students confidently choosing the ‘octet rule framework’ reasons and the ‘stable fully-filled subshell conception’ reasons within the same item and across different items (see Table 4). For example, students (5%) who confidently chose [Q3 (B2), Q4 (A1)] (oct) [Q3 (B3), Q4 (A4)] (ffs) (last entry in Table 4) decided on one of these combinations [(B1, B2, B3) only, (B2, B3) only] in item 3 as well as one of these combinations [(A1, A2, A3, A4) only, (A1, A3, A4) only, (A1, A4) only] in item 4. Tan et al. (2005b) suggested that the ‘stable fully-filled subshell’ was analogous to the ‘stable octet’, and that the latter seemed to lead ‘naturally’ to the former conception; there was no conflict between them: for example, the Na+ ion could have both a ‘stable octet’ of electrons as well as a ‘stable fully-filled 2p subshell’. Thus, the results showed that choices presenting both alternative conceptions were viable and attractive to the students, especially in items 3 and 4 where the Na+ ion was foregrounded.
Alternative conceptions | Items cross-tabulated | Percentage of students (n = 274) |
---|---|---|
Octet rule framework and stable fully-filled subshell conception | [Q1 (A2)] (oct), [Q3 (B3)] (ffs) | 5 |
[Q1 (A2)] (oct), [Q4 (A4)] (ffs) | 5 | |
[Q3 (B2)] (oct), [Q3 (B3)] (ffs) | 25 | |
[Q3 (B2)] (oct), [Q4 (A4)] (ffs) | 12 | |
[Q4 (A1)] (oct), [Q3 (B2)] (ffs) | 8 | |
[Q4 (A1)] (oct), [Q4 (A4)] (ffs) | 12 | |
[Q1 (A2), Q3 (B2)] (oct), [Q3 (B3)] (ffs) | 5 | |
[Q3 (B2), Q4 (A1)] (oct), [Q3 (B3)] (ffs) | 8 | |
[Q3 (B2), Q4 (A1)] (oct), [Q4 (A4)] (ffs) | 6 | |
[Q3 (B2)] (oct), [Q3 (B3), Q4 (A4)] (ffs) | 9 | |
[Q4 (A1)] (oct), [Q3 (B3), Q4 (A4)] (ffs) | 5 | |
[Q3 (B2), Q4 (A1)] (oct), [Q3 (B3), Q4 (A4)] (ffs) | 5 |
Cross-tabulation of students’ choices in the wIEDI showed five instances of 5% or more students consistently and confidently choosing ‘conservation of force thinking’ reasons across items 2 and 4 (see Table 2). It could be seen that the reason was not popular in item 3 (B1, 6%) compared to the ‘octet rule framework’ (B2, 36%) and ‘stable fully-filled subshell conception’ (B3, 26%) reasons in the same item (see Table 1); hence it was not consistently chosen by the students despite them being allowed to choose more than one reason in the item. Interestingly, the students who confidently chose reason A1 (cof) in item 2 preferred B2 (oct) (16%) and B3 (ffs) (11%) to the less popular B1 (cof) (4%) in item 3, but preferred A2 (cof) (24%) to A1 (oct) (9%) and A4 (ffs) (9%) in item 4 (see Table 5). The use of the phrase “a more stable system” could have favoured the use of the octet rule framework and stable fully-filled subshell conception by students over the use of the conservation of force thinking in item 3. Likewise, the focus on the removal of the second electron from sodium in item 4 could have cued the use of the conservation of force thinking rather than the octet rule framework and stable fully-filled subshell conception.
Alternative conceptions | Items cross-tabulated | Percentage of students (n = 274) |
---|---|---|
Conservation of force, octet rule framework and stable fully-filled subshell conception | [Q2 (A1), Q3 (B1)] (cof) | 4 |
[Q2 (A1)] (cof) [Q3 (B2)] (oct) | 16 | |
[Q2 (A1)] (cof) [Q3 (B3)] (ffs) | 11 | |
[Q2 (A1)] (cof) [Q3 (B2)] (oct) [Q3 (B3)] (ffs) | 11 | |
[Q2 (A1), Q4 (A2)] (cof) | 24 | |
[Q2 (A1)] (cof) [Q4 (A1)] (oct) | 9 | |
[Q2 (A1)] (cof) [Q4 (A4)] (ffs) | 9 | |
[Q2 (A1)] (cof) [Q4 (A1)] (oct) [Q4 (A4)] (ffs) | 4 |
Allowing students to write their own reasons if they believe that reasons presented to them are inadequate also enables the researchers to gain additional insights into their thinking. In item 5, 7% of the students who were confident of their comments in the ‘Other’ section for answer B indicated that sodium and magnesium had the same number of inner shell electrons; hence the shielding effects were similar, but since the nuclear charge of magnesium was higher than that of sodium, the first ionisation energy of sodium was lesser than that of magnesium. This explanation would also work for the comparison of the first ionisation energies of sodium and aluminium, as commented by the 6% who confidently chose the ‘Other’ option for answer B in item 7. In fact, cross-tabulations indicated that nine students (3%) consistently and confidently supplied the ‘same shielding effect, different nuclear charge’ explanation to both ‘Other’ sections for answer B in items 5 and 7. However, the explanation would not work for the comparison of the first ionisation energies of magnesium and aluminium in item 6, and interestingly, no student wrote this reason in the ‘Other’ section for answer B in item 6. The authors are considering the addition of these responses to the respective items, as well as others, before the release of the web-based diagnostic instrument to teachers.
It has to be noted that although the ‘I do not know’ and ‘Other’ options can also be incorporated into pen-and-paper diagnostic instruments, the students’ comments have to be manually keyed by the researchers into a software program for further collation and analysis. Google Forms greatly facilitate this process by automatically collating the responses typed in by students. It is also possible to make it compulsory for students to type in their responses if they choose the ‘I do not know’ or ‘Other’ options before they can move on to the next question. However, the researchers elected not to do so in order to make it less tedious for students so that they would be less likely to quit the quiz; besides, they could just key in any letter or symbol in the text boxes and move on.
Since the focus of this study was on the affordances of the web-based instrument rather than a survey of a population's conceptions (which was done in the previous study), the required conditions were not so stringent. Teachers informed students to attempt the quiz at their convenience after studying for it and without reference to any material. The students were also reassured that the quiz had no effect on their school grades, so it was unlikely that anyone else other than the student would attempt the quiz. It was possible that some students may ‘cheat’ by checking their textbooks or other resources, and consequently shift their understanding, allowing them to give canonical responses. However, based on the results and cross-tabulations obtained which indicated student difficulties, any effect of students referring to material during the quiz on the affordances of the web-based instrument seemed to be minimal. If the instrument is adopted as a diagnostic tool for use by a teacher, these issues do not apply as the activity becomes part of normal teaching, and the teacher can decide if students are to attempt the quiz under test conditions, or as seat work or homework.
The present study shows that a diagnostic instrument offering information about students’ understanding of a complex chemistry topic, and about students’ confidence in their understanding, can be provided in an on-line form that offers a degree of personalisation and automatic collation of responses. Teachers will likely welcome the ability to use the wIEDI, but of course the value of the wIEDI relies on teachers interpreting and applying the information it can offer to customise their classroom teaching for their particular classes.
No. | Items | Areas of alternative conceptions |
---|---|---|
1 | For Questions 1 to 4, please refer to the statement below. | |
Sodium atoms are ionised to form sodium ions as follows: | ||
Na(g) → Na+(g) + e | ||
Once the outermost electron is removed from the sodium atom, forming the sodium ion (Na+), the sodium ion will not combine with an electron to reform the sodium atom. | Octet rule framework | |
(A) True. | Meaning of electropositivity | |
(B) False. | Process of ionisation | |
(C) I do not know the answer. | Redox/displacement reactions | |
Your Response: (A) True. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) Sodium is strongly electropositive, so it only loses electrons. | ||
(2) The Na+(g) ion has a stable/noble gas configuration, so it will not gain an electron to lose its stability. | ||
(3) The process of Na(g) losing an electron is irreversible. | ||
(4) A very large amount of energy is required to add an electron to the Na+(g) ion. | ||
Other: _________________ | ||
How confident are you that your answer and reason(s) are correct? | ||
(![]() | ||
(![]() | ||
Note: Students are asked how confident they are of their reasons for every answer that they choose except for the choice of “I do not know the answer”. | ||
Your Response: (B) False. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) The positively-charged Na+(g) ion can attract a negatively-charged electron. | ||
(2) Na+(g) can react with an anion to reform the Na atom. | ||
(3) Na+(g) can undergo a displacement reaction to form Na. | ||
Other: _________________ | ||
How confident are you that your answer and reason(s) are correct? | ||
(![]() | ||
(![]() | ||
Your Response: (C) I do not know the answer. | ||
Can you elaborate on why you do not know the answer or are uncertain of the answer? _____________________ | ||
(Answer: B1) | ||
2 | When an electron is removed from the sodium atom, the attraction of the nucleus for the ‘lost’ electron will be redistributed among the remaining electrons in the sodium ion (Na+). | Conservation of force conception |
(A) True. | Nuclear attraction for an electron | |
(B) False. | ||
(C) I do not know the answer. | ||
Your Response: (A) True. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) The number of protons in the nucleus is the same but there is one less electron to attract, so each of the remaining 10 electrons will experience greater attraction by the nucleus. | ||
(2) The Na+(g) ion has one less shell compared to the Na(g) atom. | ||
(3) The total force of attraction by the nucleus is always divided equally between the total number of electrons. | ||
Other: _________________ | ||
Your Response: (B) False. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) The amount of attraction between an electron and the nucleus depends on the number of protons present in the nucleus and the distance of the electron from the nucleus. It does not depend on how many other electrons are present, although electrons do repel each other. | ||
(2) The electron which is removed will take away the attraction of the nucleus with it when it leaves the atom. | ||
(3) The attraction for the lost electron is very small and can be neglected compared to the attraction for the remaining 10 electrons. | ||
Other: _________________ | ||
Your Response: (C) I do not know the answer. | ||
Can you elaborate on why you do not know the answer or are uncertain of the answer? _____________________ | ||
(Answer: B1) | ||
3 | In an isolated system, the Na(g) atom is a more stable system than the Na+(g) ion and a free electron. | Conservation of force conception |
(A) True. | Octet rule framework | |
(B) False. | Stable fully-filled subshell conception | |
(C) I do not know the answer. | Reactions of the gaseous sodium ion | |
Your Response: (A) True. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) Energy is required to ionise the Na(g) atom to form the Na+(g) ion. | ||
(2) The Na+(g) ion has a vacant shell which can be filled by electrons from other atoms or molecules to form the Na(g) atom. | ||
(3) Na+(g) has a positive charge and will react readily with other atoms or molecules to form the Na(g) atom. | ||
Other: _________________ | ||
Your Response: (B) False. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) The average force of attraction on each electron of the Na+(g) ion is greater than that of the Na(g) atom because the attractive force of the nucleus of the Na+(g) is shared among fewer electrons. | ||
(2) The outermost shell of Na+(g) ion has achieved a stable octet/noble gas configuration. | ||
(3) All subshells are filled fully with electrons for Na+(g), whereas the Na atom has one subshell with a lone electron that gives it its instability. | ||
Other: _________________ | ||
Your Response: (C) I do not know the answer. | ||
Can you elaborate on why you do not know the answer or are uncertain of the answer? _____________________ | ||
(Answer: A1) | ||
4 | After the sodium atom is ionised (i.e. forms the Na+ ion), more energy is required to remove a second electron (i.e. the second ionisation energy is greater than the first ionisation energy) from the Na+ ion. | Conservation of force conception |
(A) True. | Octet rule framework | |
(B) False. | Stable fully-filled subshell conception | |
(C) This should not happen as the Na+(g) ion will not lose any more electrons. | Relation-based thinking – focusing on electronic repulsion and ignoring nuclear charge | |
(D) I do not know the answer. | Nuclear attraction for an electron | |
Your Response: (A) True. | Process of ionisation | |
Reason(s) (You may choose as many options as you think are correct): | ||
(1) Removal of the second electron disrupts the stable octet structure of the Na+(g) ion. | ||
(2) The same number of protons in Na+(g) attract one less electron, so the attraction for the remaining electrons is stronger. | ||
(3) The second electron is located in a shell which is closer to the nucleus. | ||
(4) It is more difficult to remove an electron from a fully filled 2p subshell. | ||
Other: _________________ | ||
Your Response: (B) False. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) The second electron is removed from a paired 2p orbital and it experiences repulsion from the other electron in the same orbital. | ||
(2) The second electron to be removed experiences shielding by the other electrons. | ||
(3) The second electron to be removed experiences a decrease in nuclear charge because some nuclear charge is removed when the first electron leaves. | ||
(4) It takes the same amount of energy to remove any electron from the same atom. | ||
Other: _________________ | ||
Your Response: (C) This should not happen as the Na+(g) ion will not lose any more electrons. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) The Na+(g) ion already has an octet of electrons in its outer shell. | ||
(2) Sodium is in Group 1 and cannot form a Na2+(g) ion. | ||
(3) The inner shell electrons of the Na+(g) ion cannot be removed as there is a very strong attraction between the positively charged ion and the remaining electrons. | ||
Other: _________________ | ||
Your Response: (D) I do not know the answer. | ||
Can you elaborate on why you do not know the answer or are uncertain of the answer? _____________________ | ||
(Answer: A3) | ||
5 | Sodium, magnesium and aluminium are in Period 3. How would you expect the first ionisation energy of sodium (1s2 2s2 2p6 3s1) to compare to that of magnesium (1s2 2s2 2p6 3s2)? | Octet rule framework |
(A) The first ionisation energy of sodium is greater than that of magnesium. | Stable fully-filled subshell conception | |
(B) The first ionisation energy of sodium is less than that of magnesium. | Relation-based thinking – the more electrons, the further away they are/filling of electrons in orbitals and ignoring nuclear charge | |
(C) I do not know the answer. | Nuclear attraction for an electron | |
Your Response: (A) The first ionisation energy of sodium is greater than that of magnesium. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) The paired electrons in the 3s orbital of magnesium experience repulsion from each other, and this effect is greater than the increase in the nuclear charge in magnesium. | ||
(2) The 3s electrons of magnesium are further from the nucleus compared to those of sodium. | ||
(3) The attractive force between the valence electron and the nucleus of sodium is stronger than the attractive force between the two valence electrons and the nucleus of magnesium. | ||
(4) Sodium has fewer electrons compared to magnesium, so the constant nuclear charge attracts the electrons more strongly in sodium. | ||
Other: _________________ | ||
Your Response: (B) The first ionisation energy of sodium is less than that of magnesium. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) Sodium will achieve a stable octet configuration if its 3s electron is removed. | ||
(2) Magnesium has a fully-filled 3s subshell which gives it stability as paired electrons are more stable and harder to remove. | ||
(3) In this situation, the effect of an increase in nuclear charge in magnesium is greater than the effect of an increase in the repulsion between its paired electrons in the 3s orbital. | ||
Other: _________________ | ||
Your Response: (C) I do not know the answer. | ||
Can you elaborate on why you do not know the answer or are uncertain of the answer? _____________________ | ||
(Answer: B3) | ||
6 | How do you expect the first ionisation energy of magnesium (1s2 2s2 2p6 3s2) to compare to that of aluminium (1s2 2s2 2p6 3s2 3p1)? | Stable fully-filled subshell conception |
(A) The first ionisation energy of magnesium is greater than that of aluminium. | Relation-based thinking – the more electrons, the further away they are/filling of electrons in orbitals and ignoring nuclear charge | |
(B) The first ionisation energy of magnesium is less than that of aluminium. | Relation-based thinking – focusing on electronic repulsion and ignoring nuclear charge | |
(C) I do not know the answer. | ‘Destruction’ of orbitals when electrons are removed | |
Your Response: (A) The first ionisation energy of magnesium is greater than that of aluminium. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) Removal of an electron will disrupt the stable completely-filled 3s subshell of magnesium. | ||
(2) The 3p electron of aluminium is further from the nucleus compared to the 3s electrons of magnesium. | ||
(3) In this situation, the effect of an increase in nuclear charge in aluminium is less than the effect of an increase in the repulsion between the electrons in its outermost shell. | ||
(4) The outermost electron of aluminium is removed from a 3p orbital, and 3p orbitals are higher in energy than 3s orbitals. | ||
(5) Removal of one electron from aluminium would result in a filled 3s subshell which is associated with greater stability. | ||
Other: _________________ | ||
Your Response: (B) The first ionisation energy of magnesium is less than that of aluminium. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) In this situation, the effect of an increase in nuclear charge in aluminium is greater than the effect of an increase in the repulsion between the electrons in its outermost shell. | ||
(2) The paired electrons in the 3s orbital of magnesium experience repulsion from each other, whereas the 3p electron of aluminium is unpaired. | ||
(3) When an electron is removed from aluminium, the 3p orbital is destroyed. When an electron is removed from magnesium, the 3s orbital is still intact. | ||
Other: _________________ | ||
Your Response: (C) I do not know the answer. | ||
Can you elaborate on why you do not know the answer or are uncertain of the answer? _____________________ | ||
(Answer: A3) | ||
7 | How do you expect the first ionisation energy of sodium (1s2 2s2 2p6 3s1) to compare to that of aluminium (1s2 2s2 2p6 3s2 3p1)? | Octet rule framework. |
(A) The first ionisation energy of sodium is greater than that of aluminium. | Stable fully-filled subshell conception | |
(B) The first ionisation energy of sodium is less than that of aluminium. | Conservation of force conception | |
(C) I do not know the answer. | Relation-based thinking – the more electrons, the further away they are/filling of electrons in orbitals and ignoring nuclear charge | |
Your Response: (A) The first ionisation energy of sodium is greater than that of aluminium. | Relation-based thinking – focusing on electronic repulsion/shielding and ignoring nuclear charge | |
Reason(s) (You may choose as many options as you think are correct): | Trend of the first ionisation energy across Period 3 | |
(1) Aluminium will attain a fully-filled 3s subshell if an electron is removed. | ||
(2) The 3p electron of aluminium experiences greater shielding from the nucleus compared to the 3s electron of sodium. | ||
(3) The 3p electron of aluminium is further away from the nucleus compared to the 3s electron of sodium. | ||
(4) The outermost electron of aluminium is removed from a 3p orbital, and 3p orbitals are higher in energy than 3s orbitals. | ||
Other: _________________ | ||
Your Response: (B) The first ionisation energy of sodium is less than that of aluminium. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) Sodium will achieve a stable octet configuration if an electron is removed. | ||
(2) In this situation, the effect of an increase in nuclear charge in aluminium is greater than the effect of an increase in the repulsion between the electrons in its outermost shell. | ||
(3) The first ionisation energy always increases across a period. | ||
Other: _________________ | ||
Your Response: (C) I do not know the answer. | ||
Can you elaborate on why you do not know the answer or are uncertain of the answer? _____________________ | ||
(Answer: B2) | ||
8 | Silicon, phosphorus and sulfur are in Period 3. How would you expect the first ionisation energy of silicon (1s2 2s2 2p6 3s2 3p2) to compare to that of phosphorus (1s2 2s2 2p6 3s2 3p3)? | Stable half-filled subshell conception |
(A) The first ionisation energy of silicon is greater than that of phosphorus. | Relation-based thinking – the more electrons, the further away they are/filling of electrons in orbitals and ignoring nuclear charge | |
(B) The first ionisation energy of silicon is less than that of phosphorus. | Relation-based thinking – focusing on electronic repulsion/shielding and ignoring nuclear charge | |
(C) I do not know the answer. | Trend of the first ionisation energy across Period 3 | |
Your Response: (A) The first ionisation energy of silicon is greater than that of phosphorus. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) Silicon has fewer electrons than phosphorus; thus its 3p electrons face less shielding. | ||
(2) The 3p electrons of phosphorus are further away from the nucleus compared to those of silicon. | ||
(3) In this situation, the effect of an increase in nuclear charge in phosphorus is less than the effect of an increase in the repulsion between the electrons in its outermost shell. | ||
Other: _________________ | ||
Your Response: (B) The first ionisation energy of silicon is less than that of phosphorus. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) The 3p subshell of phosphorus is half-filled; hence it is stable. | ||
(2) In this situation, the effect of an increase in nuclear charge in phosphorus is greater than the effect in an increase in the repulsion between its 3p electrons. | ||
(3) The first ionisation energy always increases across a period. | ||
Other: _________________ | ||
Your Response: (C) I do not know the answer. | ||
Can you elaborate on why you do not know the answer or are uncertain of the answer? _____________________ | ||
(Answer: B2) | ||
9 | How would you expect the first ionisation energy of phosphorus (1s2 2s2 2p6 3s2 3p3) to compare to that of sulfur (1s2 2s2 2p6 3s2 3p4)? | Stable half-filled subshell conception |
(A) The first ionisation energy of phosphorus is greater than that of sulfur. | Relation-based thinking – the more electrons, the further away they are/filling of electrons in orbitals and ignoring nuclear charge | |
(B) The first ionisation energy of phosphorus is less than that of sulfur. | Paired electrons in an orbital attract each other strongly | |
(C) I do not know the answer. | Trend of the first ionisation energy across Period 3 | |
Your Response: (A) The first ionisation energy of phosphorus is greater than that of sulfur. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) 3p electrons of sulfur are further away from the nucleus compared to those of phosphorus. | ||
(2) The 3p subshell of phosphorus is half-filled; hence it is stable. | ||
(3) In this situation, the effect of an increase in nuclear charge in sulfur is less than the effect of an increase in the repulsion between its 3p electrons. | ||
(4) Sulfur needs to lose one electron to have a stable half-filled 3p subshell. | ||
Other: _________________ | ||
Your Response: (B) The first ionisation energy of phosphorus is less than that of sulfur. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) More energy is required to overcome the attraction between the paired 3p electrons in sulfur. | ||
(2) In this situation, the effect of an increase in nuclear charge in sulfur is greater than the effect of an increase in the repulsion between its 3p electrons. | ||
(3) The first ionisation energy always increases across a period. | ||
Other: _________________ | ||
Your Response: (C) I do not know the answer. | ||
Can you elaborate on why you do not know the answer or are uncertain of the answer? _____________________ | ||
(Answer: A3) | ||
10 | How would you expect the first ionisation energy of silicon (1s2 2s2 2p6 3s2 3p2) to compare to that of sulfur (1s2 2s2 2p6 3s2 3p4)? | Stable half-filled subshell conception |
(A) The first ionisation energy of silicon is greater than that of sulfur. | Conservation of force conception | |
(B) The first ionisation energy of silicon is less than that of sulfur. | Relation-based thinking – the more electrons, the further away they are/filling of electrons in orbitals and ignoring nuclear charge | |
(C) I do not know the answer. | Trend of the first ionisation energy across Period 3 | |
Your Response: (A) The first ionisation energy of silicon is greater than that of sulfur. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) Sulfur will have a stable half-filled 3p subshell if an electron is removed. | ||
(2) The 3p electrons of sulfur are further away from the nucleus compared to those of silicon. | ||
(3) In this situation, the effect of an increase in nuclear charge in sulfur is less than the effect of an increase in the repulsion between its 3p electrons. | ||
(4) The silicon atom has fewer electrons in its outer shell to share the nuclear attraction. | ||
Other: _________________ | ||
Your Response: (B) The first ionisation energy of silicon is less than that of sulfur. | ||
Reason(s) (You may choose as many options as you think are correct): | ||
(1) In this situation, the effect of an increase in nuclear charge in sulfur is greater than the effect of an increase in the repulsion between its 3p electrons. | ||
(2) The first ionisation energy always increases across a period. | ||
Other: _________________ | ||
Your Response: (C) I do not know the answer. | ||
Can you elaborate on why you do not know the answer or are uncertain of the answer? _____________________ | ||
(Answer: B1) |
This journal is © The Royal Society of Chemistry 2019 |